Correction of the effect of spatial gradient field distortions in diffusion-weighted imaging

Abstract

A general mathematical framework is formulated to characterize the contribution of gradient non-uniformities to diffusion tensor imaging in MRI. Based on a model expansion, the actual gradient field is approximated and employed, after elimination of geometric distortions, for predicting and correcting the errors in diffusion encoding. Prior to corrections, experiments clearly reveal marked deviations of the calculated diffusivity for fields of view generally used in diffusion experiments. These deviations are most significant with greater distance from the magnet's isocenter. For a FOV of 25 cm the resultant errors in absolute diffusivity can range from approximately −10 to +20 percent. Within the same field of view, the diffusion-encoding direction and the orientation of the calculated eigenvectors can be significantly altered if the perturbations by the gradient non-uniformities are not considered. With the proposed correction scheme most of the errors introduced by gradient non-uniformities can be removed.